1. Field of the Invention
The present invention relates to interconnecting in series a sequence of semiconductor devices and, more particularly, to interconnecting in series an array of photovoltaic devices.
Solar cells represent a source of electrical energy based on an inexhaustable "fuel", with the operation of such devices being nonpolluting. The primary difficulty in the use of a photovoltaic device based electrical energy source has been economics. The costs of fabricating solar cells have heretofore prevented widespread use of such cells for providing electrical energy, and have confined such use to special situations where the fabrication economics do not make the use thereof prohibitive.
A major improvement in economics follows from the use of amorphous silicon as the semiconductor material in the solar cells rather than crystalline silicon. Crystalline silicon is an indirect-band-gap material meaning that a lattice phonon is required to participate in the absorption process with an incident photon. Thus, crystalline silicon absorbs electromagnetic radiation relatively weakly. Amorphous silicon, on the other hand, is a direct-band-gap material of an effectively larger bandgap in which the incident photon can be absorbed without an interaction being required of lattice phonons. As a result, an amorphous silicon layer of a given thickness can absorb as much electromagnetic radiation from the sun as can a crystalline silicon layer many times its thickness, typically in a thickness ratio of fifty to one, even though at a somewhat shorter wavelength range. Thus, very much thinner films of amorphous silicon can be used and still absorb the same amount of incident radiation energy, a structure which reduces the cost of a solar cell considerably.
The use of amorphous silicon, however, has problems of its own. The pure material has a low resistivity and is insensitive to the addition of doping impurities because there are relatively large numbers of electronic energy states occurring at energy values that would be in approximately an energy state gap in crystalline silicon. Thus, this region of amorphous silicon is often referred to as a "pseudogap" and is located in the mobility gap between extended energy states. These additional electronic states arise because of the presence of small voids throughout the amorphous silicon which give rise to various dangling bonds and distorted bonds between silicon atoms.
This situation, which would otherwise make amorphous silicon a poor candidate for forming solar cells, is greatly improved by introducing a substantial concentration of hydrogen into the amorphous silicon, usually to the extent that hydrogen represents many atomic percent of the resulting material. This hydrogenated amorphous silicon is usually designated as a-Si:H. This improvement follows from hydrogen forming bond with the silicon to eliminate dangling bonds, and also breaking distorted bonds, through the hydrogen bonding to the silicon atoms. These effects, and others, lead to material which has a relatively well defined energy gap and in which the semiconductor properties can be controlled by the doping of further impurities. That is, n-type conductivity material can be provided by doping with phosphorus, and p-type conductivity material can be provided through doping with boron, as examples. This situation permits the forming of p-n junction structures or p-i-n structures ("i" meaning intrinsic or near intrinsic semiconductor material) so that amorphous silicon structures subject to incident electromagnetic radiation can be operated as photovoltaic solar cells.
Such solar cells are usually formed in a large array of individual cells to capture large amounts of incident sunlight. However, because p-n junctions or p-i-n layer arrangements formed in doped a-Si:H yield photovoltaic cells with open circuit voltages measuring several tenths of a volt, there is a desire to electrically interconnect at least some cells in the array in series to provide a greater output voltage. Typically, such cells are formed as a "sandwich-like" structure on a substrate with such cells having, as a general matter, two conductive layers with a semiconductor material layer therebetween where one of the conductive layers is directly on the substrate. The semiconductor layer has a p-n junction or p-i-n layer arrangement therein more or less parallel to the conductive layers. One of the conductive layers is transparent to pass incident electromagnetic radiation to the semiconductor material layer (the substrate will also be transparent if it directly supports the transparent conductive layer). There is difficulty with this arrangement in electrically interconnecting the conductive layer adjacent the substrate in one cell, because of it being covered by the other "sandwich" layers, to the conductive layer of an adjacent cell on the opposite side of the semiconductor material layer therein.
An arrangement is desired for effecting such interconnections which is economical and reliable. Such interconnections must be made in a process compatible with fabricating large volumes of solar cells.